David Erickson posted an entry I meant to write for years. Open source is great at a lot of things, but let's face it, documentation is often incredibly lousy. I'm spending a lot of times these days looking at what we can and can't do with our PHP/Wordpress/OpenAds toolset. Sometimes it really feels like a root canal to get answers to seemingly simple questions. MSDN is far from perfect, but it's heaven compared to, say, how variable scoping is defined on the php.net site, empty pages in the Wordpress codex, or confusing PAN/MMM roadmaps (though OpenAds is at last starting to clean its house on that front).

For-profit companies built around open source projects need to understand this is a critical missing piece. Unlike David, I don't have high hopes that users are suddenly going to turn into good documentation writers. This is a tedious, time-consuming job and you need to keep the content updated. It has to be funded because it won't work as a labor of love. Yes, you can buy O'Reilly books, but I don't think the book market alone solves it.

(Sorry, my comments are broken and won't be fixed for a while, please comment on your own blogs!)

Somewhat related from the archives: Open Companies More Important than Open Source.

02/02/08 update: Bottom line, all weblog apps suck in some way.

Are you looking forward to the next release of Microsoft Office, whenever that is and whatever it will do? Yeah, me neither. However, I can see how a lot of value could be delivered on top of supposedly ultra-mature products. Consider the needs of a virtual team doing things such as product management and business development. Sure, there's an increasing number of web apps to cover such needs, but frankly, you can Ajax them to death and Excel's frontend is still going to beat them to a pulp in terms of responsiveness, flexibility and depth (yes I want to be able to sort and filter and use garish colors if I want to). I tried Jot during its beta and even the regular text wiki experience was painful; every web grid I've played with was just that, a toy. Imagine that enabling keyboard shortcuts in GMail was considered a breakthrough. (Breathing down the neck of MS-DOS if you ask me, but everything old is new these days.)

What I really want is seamless integration between Excel on the frontend and an online relational database on the backend. No, I DON'T want some intranet Office Server crap or Groove client overload that I have to install and maintain myself, because that's just not going to happen. And I'm not talking about just connecting Excel to some SQL source and browse it remotely (a feature I've been using once in a while), because this starts from the assumption someone created a SQL database in the first place. Nice but not nearly enough.

What I want is an internet application with a desktop frontend, with a choice of providers you can plug into, just like you can source other hosting services. The whole "internet Excel stack" should automagically normalize and synchronize the pseudo database work that most people do with it. And if it looks and tastes like a bunch of names and addresses, I should likewise be able to read/write/synch them through Outlook contacts. Wikify/blogify Outlook Today to have a mini-portal to point people to stuff and keep them on the same page (putthat stuff on the private web too), and we're all set.

The magic part so unlike what's available today is that you could start right from Excel, doing the junk you've done with it for a decade, but now it gets all cleaned up and consolidated so that your team can jump in without email attachments and tedious manual mergers. With synchronization down to the cell level (Just Figure It Out, sharing whole files is technology from last century and it's a waste of our time).

Recap in case someone from Microsoft reads this (Scoble ping):

• ASP service as a priority, not a half-baked afterthought.


• No extra client install on top of the latest default MS Office install.


• Make it as much backwards compatible as possible.


• No crap that just doesn't happen to work on Office Mac.


• No glue grunt work necessary on my end to make the magic happen.


• Data entry starts with Excel and Outlook.


• Data goes into tidy, secure, backed-up server heaven by wizard magic.


• Integrated basic wiki/blog/portal.


• No butterfly, no dog, no silly animal whatsoever.

Or even shorter: no server, no install, no programming. Get back to your end-user roots, drop the full-time IT department requirement mindset, and charge a tax to your product and marketing teams every time they use the word "deploy". Whie you're at it, drop silly names such as "Microsoft Office SharePoint Portal Server 2003" (I'm not making this up), really this should be how Office works out of the box. Do the hard work so I don't have to, damn it! A team of several people needs to be able to be up and running almost instantly: "Got Excel? Fine, click on this URL, here's your password, and start updating the sales pipeline already."

Of course it's not going to happen. Microsoft will die fat through death by incrementalitis (propagated by the nasty Enterprise virus). For people who start companies now, Microsoft is the new IBM. If you grow your business to the point where you need some IT guys, you're going to whip them until they figure out "the internet way" rather than deploy heavy infrastructure within your walls.

Update: Robert tells me I might be surprised by what's coming. We'll see, I'd love to be proven wrong.

08/29/05 update: a comment below gets me to look again into Salesforce.com, which leads me to Salesforce.com: The MS Office Killer, Salesforce.com Vs. Microsoft Office? No contest. This Multiforce page is a bit abstract though.

09/19/05 update: Server-side Excel (different idea, interesting too).

10/10/05 update: Excel 12 blog.

06/06/06 update: Google Spreadsheets complements Excel?.

08/04/15 update: Zoho Sheet integration into Zoho CRM.

As recently as 18 months ago it seemed there wasn't much to handle software product development except either bug databases, generic wikis and intranet platforms, or maybe expensive and (for my purposes) cumbersome enterprise software, which is not what I'm interested about.

Now that I'm looking again at this though (we need something to keep our growing team in synch at Soflow), there are at least three new applications meant to address the need for the right mix of structure and flexibility:

• Basecamp


• Rally (they raised $4.5M recently)


• Trac

Atlassian provides both issue tracking and collaboration but I'm looking for a single tool that seamlessly supports the whole product management and development process, from business case to user requirements to functional specs to tracking development progress, tests and debugging. Borland has CaliberRM but it needs its own server (the fact it comes both with a web and a desktop client is attractive though) and is not exactly cheap. But if I look in that direction there are plenty of requirement management tools.

Here are my notes so far: